| --> |
At CallCenterOps.com we’re dedicated to providing information about operations management to those involved in real-time customer service via call centers.
Learn how to advertise on this site. |
|
|||||||
| General Discussion The CallCenterOps Forum allows you to seek the advice of other knowledgeable call center professionals. Post your call center related question and contribute your opinion to others seeking advice. (No advertising is accepted - posts will be removed.) |
![]() |
|
|
LinkBack | Thread Tools | Display Modes |
|
|||
|
Forecast accuracy metrics
I do forecasting (2 weeks to 2 months out) for our contact centers. I am trying to find some benchmark data for how forecast accuracy is measured in other call centers. Could I ask other forecasters to give me a description of the performance measures you are held to?
My measurement is fairly complex, but is attempting to hold me accountable for each interval of the day rather than EOD results where I could do poorly in the AM, but make up for it in the afternoon. Attempting to achieve a consistent customer experience. At the same time trying to place the most weight on the intervals that have the most impact. So being 100% off at 2am when we get 2 calls instead of the 1 forecast, carries little to no impact where a 7% deviation at our 800am volume peak is very impacting. I am measured on the weighted percentage of intervals achieved within a certain range for a day for 3 metrics: Volume (55% of the rating), AHT (35% of the rating) and Requireds (10% of the rating). One of our 24 hours skills would have 48 interval opportunities to achieve a certain level of deviation. I measure them first at the absolute deviation level by interval (for example: forecast vol 100/actual vol 91 = 9% deviation) Then the ultimate rating is based on that deviation weighted by the volume it represents to the whole day. (same example; that 91 calls represents .21% of the 3864 volume we got for the day) for this skill I have the level of deviation set at .35% or less to achieve the interval. This .35% is based a careful analysis of how much requireds increase/decrease, basically how much fluctuation can the skill absorb. The 9% absolute or .21% weighted deviation would counted as achieved. This example interval was not during our peak time, but earlier in the AM. All of the achieved intervals are counted up and measured against the opportunities for the day. Following this example; I had 48 opportunities for the day and I achieved 43 of them for a 90% rating for the day for Volume. I had 48 opportunies for AHT and I achieved 42 of them for an 88% rating and for requireds, I also had 48 opportunities and I achieved 41 of them for 85%. My overall rating for the day for this skill would be: 90% vol x .55% weight + 88% AHT x .35% + 85% Req x .10% = 89% overall rating for the day for this skill. -75% in poor 75-80% is below average 80-85% is average 85-90% is above average 90%+ is exceptional Wwwhhhheeeewwww....I told you it was complex! It's even a bit more complex as all of our various skills ratings are weighted and rolled up to 1 end result for my goal. For this same day, the end of day results were: .21% absolute deviation on volume, 2.14% for AHT and .55% for requireds. I have always heard that +/-5% End of Day deviation is industry standard, but I can't find and documentation of that. Also how many skills do you forecast for? I have a total of 46 skills that I forecast for each week. |
![]() |
| Thread Tools | |
| Display Modes | |
|
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Email Support - Service Level Metrics | ltaylor | General Discussion | 3 | 11-07-2008 03:41 PM |
| Looking for metrics but need to clarify what... | reemusk | General Discussion | 0 | 03-03-2006 03:29 AM |
| Does anyone here Forecast.... | jp2time | General Discussion | 1 | 01-21-2006 07:15 PM |
| Are there any tools to forecast calls ONLY, not workforce? | Jose | General Discussion | 1 | 09-21-2005 07:04 AM |